NotEqual / Operator Layer

Compare values for inequality between a column and a constant, or between two columns. Similar to pandas df['A'] != df['B'] or df['A'] != 5, numpy.not_equal(), or R's != operator. Returns a boolean column (true where values differ).

Mathematical form: $y_{i} = {t r u e f a l se if x_{i} \neq = v otherwise$

Common applications:

Anomaly detection (value != expected)
Change detection (current != previous)
Error checking (status != success)
Data cleaning (remove unwanted values)
Filtering outliers
Exception handling

Note: Consider precision issues when comparing floating-point values.

Table

SelectLeft

column

The primary column for comparison. Forms the left side of the inequality check (e.g., current value, test measurement, input data).

SelectRight

oneof

Integer

Value

i64

64-bit signed integer for inequality comparison. Range: -2^63 to 2^63-1. Use cases:

Error detection (code != 0)
State changes (status != -1)
Boundary checks (value != limit)
Flag verification (setting != default)

UnsignedInteger

Value

u64

64-bit unsigned integer for inequality comparison. Range: 0 to 2^64-1. Examples:

Capacity checks (size != maximum)
Counter validation (retries != limit)
ID verification (user_id != root_id)
Resource monitoring (available != 0)

Float

Value

f64

64-bit floating-point for inequality comparison. Examples:

Threshold checking (temperature != 0.0)
Calibration verification (reading != baseline)
Signal processing (amplitude != noise_floor)
Rate monitoring (growth != stagnant)

Column

Other

column

The second column for inequality comparison. Common pairs:

measured != control values
new_data != old_data
sensor_1 != sensor_2 (redundancy check)
actual != predicted (model evaluation)

PropagateNull

bool

true

Controls null value handling:

true: null != null returns false, null != value returns null
false: null != null returns null, null != value returns null

Default is true for standard SQL-like null comparison behavior

AsColumn

name

Name for the new column. If not provided, the system generates a unique name. If AsColumn matches an existing column, the existing column is replaced. The name should follow valid column naming conventions.